Preflight medical screening of patients

Travel by commercial jet aircraft is a popular, expeditious, safe and generally affordable mode of transportation. Millions of passengers fly all over the world every year. A proportion of them are disabled, or have experienced recent injury or illness. Despite the comfort and convenience of modern commercial aircraft, there are environmental factors that may affect some persons, particularly those with limited physiological reserve. By far the most important one is reduced barometric pressure (PB), which may expose patients with altered pulmonary gas exchange to a poorly tolerated hypoxic stress. Altitude is associated with a decrease in PB. The cruising altitude of most commercial airliners is 9,000±12,000 m, which corresponds to a PB decrease from the sea-level 101 kPa (760 mmHg) to 30.7±19.3 kPa (231±145 mmHg). The inspired partial pressure of oxygen (PI,O2) at 9,000 m can be calculated from the equation: PI;O2  FI;O2PBÿPair;H2O378C  5:1 kPa 39 mmHg; 1


1
where FI,O 2 is the inspiratory oxygen fraction (0.21, which does not change with altitude) and Pair,H 2 O(378C) the partial pressure of water at 378C (6.3 kPa (47 mmHg)). This is probably the extreme PI,O 2 compatible with human life, tolerable only for some highly trained mountaineers climbing Mount Everest (8,840 m) after several weeks' acclimatization [1]. Acutely, exposure to an altitude of 9,000 m is associated with loss of consciousness in <1 min [1].
Hypobaric hypoxia, in aeroplanes, is prevented either by increasing the FI,O 2 , with an~1% increase corresponding to a decrease in altitude of 300 m, or by increasing cabin pressure. Aviation regulations specify that flights in nonpressurized aircraft at an altitude of >3,000 m require supplemental oxygen delivered by a face mask, and that aircraft cruising at altitudes of >6,000 m must be pressurized [2]. It is technically difficult, and costly, to pressurize cabins equivalent to sea-level altitudes. The cabin pressure used is a compromise between the need to avoid any untoward physiological effect of hypobaric hypoxia and technical and economic imperatives [3]. The American Federal Aeronautical Regulations specify that the cabin pressure must not be lower than an equivalent altitude of 2,434 m, with tolerance to a pressure equivalent to a cabin altitude of up to 3,000 m in case of a technical problem [2]. Supplemental oxygen is mandatory at pres-sures equivalent to higher cabin altitudes, as it is in unpressurized aircraft flying higher than 3,000 m [2].
In 1988, COTTRELL [4] reported in-flight cabin pressure measurements relative to altitude on 204 regularly scheduled commercial aircraft flights. He showed a median pressure exposure for all flights equivalent to an altitude of 1,894 m, with a range from sea level to 2,717 m, and expressed concern that higher flying newer design aircraft together with economic pressure to limit costs might lead to cabin pressures equivalent to higher altitudes. This concern has not been confirmed. The present author measured cabin pressures equivalent to altitudes ranging 580±2,460 m with a wristwatch altimeter during 28 European and trans-Atlantic flights in 1999.
At a pressure equivalent to a cabin altitude of 2,438 m, the PI,O 2 is decreased by 28% from that at sea level, corresponding to an FI,O 2 decreased to 16%. Approximately fifty years ago, DRIPPS and COMROE [5] marvelled at how little and late cardiovascular and respiratory adaptations to acutely decreased PI,O 2 were to be observed in normal subjects. These authors showed that, when the FI,O 2 is progressively reduced, there is hardly any increase in ventilation down to an FI,O 2 of~0.12, whereas cardiac frequency starts to increase at an FI,O 2 of~0.16±0.18, leading to a 15% increase at an FI,O 2 of 0.12 [5]. In normal subjects breathing an FI,O 2 of 0.125, arterial oxygen tension (Pa,O 2 ) falls to~5.3 kPa (40 mmHg), there is no significant change in ventilation and oxygen delivery to the tissues is maintained by means of an increase in cardiac output of~20% [6]. It is of interest that a Pa,O 2 acutely brought down to 5.3 kPa is not associated with a significant change in blood pressure [5,6]. It is thus evident that the lowest PI,O 2 permitted in the commercial aircraft cabin cannot be considered to cause cardiorespiratory stress in normal subjects at rest. At exercise, the same level of inspiratory hypoxia is associated with a decrease in maximum oxygen consumption, but is otherwise normally well tolerated [1]. During the l9th Olympic Games held in Mexico city, at an altitude of 2,240 m, the mean maximum speed during long-distance running contests was reduced by~10%, in proportion to the decreased maximum oxygen consumption, but there was no report of any abnormal increase in the incidence of medical problems.
The story may, however, be different in patients with pre-existing cardiac or pulmonary diseases. These patients may be exposed to excessive in-flight hypoxaemia, as defined by a Pa,O 2 of <7.3 kPa (<55 mmHg) by the Aerospace Medical Association [7], or by a Pa,O 2 of <6.7 kPa (<50 mmHg) by the American Thoracic Society [8]. How predictable is in-flight hypoxaemia? It could be imagined that cabin pressures could be transposed into the alveolar gas equation, and in-flight Pa,O 2 estimated from the preflight alveolar to arterial oxygen tension gradient. However, the ventilatory as well as the cardiac responses to hypoxaemia may be reduced to variable extents, in relation to more or less compromised lung and chest wall mechanics, altered respiratory centre function due to chronic hypercapnia exposure and abnormal ventricular function. Conversely, although diffusion limitation may be unlikely, changes in PI,O 2 , ventilation and cardiac output all are prone to affecting ventilation/perfusion (V '/Q ') distributions and the resultant Pa,O 2 , depending on the type and extent of baseline alterations in gas exchange [9].
It is thus understandable that several groups have sought to predict in-flight Pa,O 2 by means of preflight testing at low oxygen tensions. This has been done either by reducing the FI,O 2 , connecting the patient to a low-oxygen circuit [10±12], or by reducing PB, placing the patient in a hypobaric chamber [13±15]. There is no physiological reason to believe that the cardiorespiratory effects of hypobaric or normobaric hypoxia would be different [16]. Accordingly, both types of hypoxia testing have produced comparable results [17].
DILLARD et al.  [7,10,18]. This opinion has been recently challenged by CHRIS-TENSEN et al. [19]. These authors measured arterial blood gas levels at sea level, and at PB equivalent to altitudes of 2,438 and 3,048 m in a hypobaric chamber, both at rest and during light exercise, in 15 patients with COPD and a sea-level Pa,O 2 of >9.3 kPa (>70 mmHg). At 2,438 m, five of the patients had a Pa,O 2 of <6.7 kPa (<50 mmHg), and this hypoxaemia was aggravated during mild exercise at a level corresponding to a quiet in-flight walk. It is of interest that the preflight aerobic capacity of the COPD patients was well correlated to Pa,O 2 at pressures equivalent to 2,438 m. All of the patients with a preflight aerobic capacity of >12.1 mL . min -1 . kg -1 maintained an in-flight Pa,O 2 of >6.7 kPa at pressures equivalent to 2,438 m. Thus simple noninvasive exercise testing might perhaps suffice for the prediction of good tolerance to in-flight hypobaric hypoxia in COPD patients. Until now, exercise testing has not been included in guidelines for the preflight medical screening of patients [7,8,20].
In this issue of the European Respiratory Journal, ROBSON et al. [21] report their experience with hypoxia inhalation testing and an interpretational algorithm for the identification of patients at risk of respiratory distress during commercial air travel. These authors studied 28 patients with respiratory disease. Four of them, with baseline oxygen saturation measured by pulse oximetry (Sa,O 2 ) of <90%, were considered unfit for air travel. The rest were submitted to hypoxic breathing with an FI,O 2 of 0.15, and their arterial blood gas levels were measured when the Sa,O 2 decreased to <90%. Two patients showed a Pa,O 2 of <6.7 kPa (<50 mmHg) during the hypoxia test (5.3 and 6.0 kPa (40 and 45 mmHg), respectively), which was corrected by supplemental oxygen. In-flight supplemental oxygen was prescribed for these two patients. The patients allowed to fly, and also one of the patients who flew against medical advice without supplemental oxygen, did not report any in-flight problems.
What is the clinical relevance of preflight hypoxia testing? Preflight normobaric or hypobaric hypoxia testing has not been associated with any reported symptomatic intolerance, even in those patients whose Pa,O 2 acutely decreased to <6.7 kPa (<50 mmHg) [10±15, 17,19,21,22]. In a study performed under more realistic conditions, SCHWARZ et al. [23], measured arterial blood gas levels in 13 patients with COPD during flights in unpressurized aeroplanes at altitudes of 1,650 and 2,250 m. None of the patients reported any symptoms, in spite of rather spectacular decreases in Pa,O 2 , from a mean preflight value of 9.0 kPa (68 mmHg) (range 7.3±10.5 kPa (55±79 mmHg)) to 6.8 kPa (51 mmHg) (range 4.3±8.1 kPa (32±61 mmHg)) at 1,650 m and 6.0 kPa (45 mmHg) (range 3.3±7.4 kPa (25± 56 mmHg)) at 2,250 m [23]. In an invasive study on the variability of pulmonary vasoreactivity to hypoxia, WEIT-ZENBLUM et al. [24],observed the effects of low oxygen levels in 28 patients with COPD. Pa,O 2 decreased from 8.00.80 kPa (606 mmHg) (meanSD) at baseline, to 5.70.93 kPa (437 mmHg) at an FI,O 2 of 0.15 and 5.10.67 kPa (385 mmHg) at an FI,O 2 of 0.13. Again, these impressive decreases in Pa,O 2 were well tolerated. It has long been known that COPD patients may experience transient episodes of severe hypoxaemia, with Pa,O 2 as low as 3.3±4.0 kPa (25±30 mmHg) without immediate harm [25]. Dangerous levels of hypoxaemia have never been defined. Severe acute hypoxaemia carries risks of cerebral dysfunction, myocardial ischaemia and lung oedema, but, until now, there has been no means of determining the Pa,O 2 below which a patient would be in danger, especially in cases of pre-existent chronic hypoxaemia.
Preflight hypoxia testing is cumbersome, expensive, time consuming and not generally available. The opinions of respiratory physicians regarding assessment methods and criteria for recommending in-flight oxygen seem to vary widely [26]. Guidelines produced by the Aerospace Medical Association [7], the American Thoracic Society [8] and the European Respiratory Society [20] describe or mention predictive equations or preflight hypoxia testing, but do not recommend their routine clinical use. In the meantime, the incidence of significant in-flight respiratory problems seems to be extremely low, in the range of one respiratory incident resulting in flight diversion per 5±10 million passengers [26,27].
When, then, is it necessary to prescribe in-flight supplemental oxygen? It should certainly be prescribed to the pulmonary patient who already receives continuous or intermittent oxygen at home. Supplemental oxygen may also be considered in the hypoxaemic patient with a markedly decreased FEV1. In-flight oxygen is not indicated for cardiac patients, who most generally tolerate altitude exposure very well [28]. An exception may be the patient with pulmonary hypertension, potentially aggravated with mild degrees of additional hypoxic pulmonary vasoconstriction. If in doubt, it is important to note that in-flight oxygen is probably less expensive than preflight hypoxia testing.
After more than a quarter of a century of preflight hypoxia testing, it is now up to its proponents to design a clinical study showing that this practice is clinically relevant and cost-effective.